Despite significant improvements in the treatment of primary cancer, successful treatment of metastatic disease remains a challenge. Indeed, the 5-year survival rate for patients with tumors distant from the primary site has not significantly improved in the past decade (1). The search for an alternative to systemic chemotherapy has thus been a priority, especially in the context of metastases.Prodrug/enzyme treatments have been used to circumvent some side-effects of chemotherapy (2, 3). One of the enzymes that has met with some success is bacterial cytosine deaminase (bCD) from Escherichia coli. bCD converts the non-toxic prodrug 5-fluorocytosine (5-FC) to 5-fluorouracil (5-FU) (4). The latter potent chemotherapeutic agent has been used for decades in therapy of colorectal, head and neck, and breast cancer, among others. It is, however, associated with debilitating side-effects such as cardiotoxicity, diarrhea and myelosuppression (5) since high doses of systemic 5-FU are required to achieve significant tumor concentrations. To minimize sideeffects of 5-FU, targeted delivery of bCD to tumor cells coupled with systemic administration of 5-FC is also being considered. Monoclonal antibodies conjugated to the enzyme, as well as viral vectors and non-viral vectors, have been used to achieve this goal (6-13). Recent studies have demonstrated prostate cancer cell-specific nanoparticle delivery of bCD to tumor xenografts by specific targeting of prostate-specific membrane antigen that resulted in improved tumor control (13,14) following localized conversion of 5-FC to 5-FU.However, many types of cancer do not express cancer cell-specific surface receptors or antigens, making it important to determine the effect of 5-FU formed from bCD in the absence of targeted delivery. The therapeutic effect of bCD/5-FC in a metastatic disease setting remains sparsely evaluated (15-17). Our purpose in this study was to evaluate bCD/5-FC treatment in the metastatic setting.Several mutants have been created for an enhanced conversion rate of 5-FC to . A triple-mutant (bCD 1525 ) displaying high specificity for 5-FC (21) exhibited increased antitumor activity and also bystander effect, a phenomenon by which the cytotoxic metabolite is transferred to nearby cells via gap junction intercellular communications (22, 23). Here we used the bCD 1525 triple-mutant enzyme to 2195